Ocado Group plc (OCDO) Earnings Call Transcript & Summary

Thank you very much for joining us this afternoon. We've got 1.5 hours. We've got a hard stop at 4:30. So we're going to try and rattle through the slides in about 45 minutes if we can, leaving the balance for Q&A. I think we're just 2 weeks short exactly to the day on 6 months ago of when we did our modelling seminar from an EBITDA perspective on the 25th of May. And we've heard a lot of feedback subsequently that we really enjoyed that session, but we'd like to know a lot more about cash flow, returns on cash flow because that's particularly important to the Ocado business model and the equity narrative, the investment case for Ocado Group. So what we're going to try and do in the next 1.5 hours, the next 45 minutes. Next slide, please. Sorry, I clicked through, I don't have [indiscernible] myself there. So 3 core objectives. First of all, a better understanding of the key drivers of cash flows, the unit economics and then the drivers that move those economics in either an up or down direction. Providing a framework there for you to assess how for each of these businesses, the 3 business models that we highlighted 6 months ago, might develop over the next 4 to 6 years. Again, we're using the midterm as the sort of the narrative of the description of that 4- to 6-year horizon. We're going to take a deeper dive in particular into the cash flows in the technology Solutions business. That's the key future value driver for the group, and it's where I know that most of you are most interested when we think about Ocado Group in the 3 different business models. What hopefully the session will do will reinforce the conviction that we have is that our business is fully funded, on the road to cash flow positive in the midterm. And we can demonstrate the market-leading returns that we believe our capital will produce as we invest in our automated warehouse model. Okay. Those are the objectives. Here's the agenda. But first of all, just going to set the scene just a couple of slides. As a reminder of the continuous improvement journey that we have been on and remain on around the deployment of our capital, bringing down capital intensity and improving productivity at the same time, showing the journey. We're going to review our underlying business models, the 3 businesses that we highlighted 6 months ago. We're going to go through the drivers of each, and then we're going to bring it all together. What does that mean? So a track record of continuous improvement. What we wanted to show here was how capital intensity and units per hour productivity have evolved since the first warehouse went live in 2002 in Hatfield over the last 20 years. And we don't have precise numbers around the CapEx on Hatfield, but we're pretty sure it's around that 20% of sales capital intensity. Similarly, on units per hour, you can see pretty good sense of this one in the 100s, and in fact, still in the 100s, productivity in Hatfield. Today's model, the Ocado Smart Platform model can do 4 weekly shopping baskets in less than 1 hour. Over 200 units per hour, we're picking ready for delivery to our customers and match productivity. We are moving towards that number being well in excess of 300 and heading towards 400. Our ultimate goal, as I've said this before, and Tim describes it as a lights-out CFC, fully automated. No people in the warehouse at all, on-grid robotic pick, doing a very good percentage of the range of that right now in Purfleet, and we'll see the numbers later. And then also frame load our part of that journey, but it's ongoing. At the same time, bringing down our capital intensity. Most recently, we've highlighted what our capital intensity is as a proportion of sales. We're going to show where we are today and where we are getting to in our Re:Imagined innovations that we highlighted in January. And then the ultimate goal, again, is to bring capital intensity down even further to the 10% gross level. More on that later. Whilst we've introduced or reduced capital intensity and we've increased productivity, at the same time, we have improved the cash flow phasing for ourselves and for our partners, really important point. And you'll see a detail on this, particularly important when it comes to payback and returns on capital. Pre 2016, the U.K. legacy sites of Hatfield and Dordon, typically, we would invest 100% of capital upfront went on live -- and when we went live. Under our new modular site design, around 65% of the CapEx is invested in tranches as modules, which are basically units of sales capacity, are added to a site once it's built. Typically, we go live with 2 modules, and we add as we go along, and therefore, the CapEx is being incurred as we go along rather than all upfront. We face that CapEx to the ramp profile. We incur CapEx upfront, the balance 60 -- 35% of the CapEx when we put down the slab in the site before adding the hive on the bots and so on. There is a degree of fixed and variable. We'll go into that detail shortly. The return on capital employed that we're getting today is 22%. That's the perfect example that we gave 6 months ago. All of the fixed CapEx, as we see shortly, is covered by the 4% upfront fee as a proportion of sales that we get from our partners. So in other words, we are investing a net zero cash outflow upfront in the fixed assets in a CFC before then incurring the module CapEx as we roll out modules. We'll see shortly that the module rollout CapEx has a payback of less than 2 years. So I want to be explicit about that. And we will also show that the peak cash outflow under this model is now just GBP 15 million per warehouse, having previously been double that amount of GBP 30 million. That's a good news story because it's less cash outflow before the return starts to follow. Going forward, Re:Imagined basis from 2024 onwards, we're expecting a higher return on capital employed. If you do the math, you actually get to a 50% return on capital employed. I think right now, we're saying 40% plus at that stage as we mature into that return target. For Ocado, it means lower CapEx and higher fees. We are going to be explicit now about what that means for the client fees. And at the same time, for our clients, they'll see a material reduction in CapEx and operating costs, and we'll show that as well on the components. All of these are the enabler to the path to a net zero cash outflow for the upfront investment in the material handling equipment in the CFC. So reducing those peak cash outflows, improving time lines to pay back for our clients for Ocado Group. A reminder again, I'll whizz through these next couple of slides. On the left are the way we have or currently report our group numbers, Ocado Retail Solutions and Logistics International Solutions and on the right is the model that we plan to adopt. The straight line here is an important one. I'm happy today to confirm that we are on track to report under the new basis in fiscal '23. Six months ago, I was asked the question, I think I pretty much hedged my bets on that one. We believe we've got there now. I think most importantly, though, is that we'll be constructing our KPIs around those 3 business models and around the EBITDA numbers and the cash flow numbers, the EBITDA number which you saw 6 months ago, cash flow you're going to see today. So our KPIs will be drawn from those economics, you can track the progress that we're making. I won't go through this. This is something you can read at your leisure. It's just being a bit more explicit about exactly what the inflows and outflows are from a cash perspective, under the 3 different business models. A lot of good detail in here, but much of it is covered in the slides that are going to follow. So Ocado Retail. Just a few words on Ocado Retail. It remains the best example for us of the retail model and what the retail model can achieve. It's very important for us, not just for that but also as a test ground for new innovations close to home in the U.K. The key messages for Ocado Retail that we're sharing with you this afternoon. We're confident of a return to high mid-single-digit EBITDA. We're going to show you the reasons why. And we believe there's further potential margin expansion to around 7% with the Re:Imagined innovations. The capacity already installed and invested, we're going to show how that delivers by the midterm revenues of GBP 3.9 billion. Important point to make here in May, you will have seen GBP 4.5 billion for that number. We're going to highlight shortly on a slide, and I'll reference it now. We've taken the decision. We think it's a sensible decision to pause the Northwest and the Southeast CFCs that we highlighted for opening in '24 and '25, respectively, which is why the GBP 4.5 billion is now GBP 3.9 billion because of that distinction. That may change. It's a pause. It's not a stop, but we think it's a sensible thing to do given the surplus capacity that we have today. Notwithstanding that, at that 7% margin that will deliver EBITDA in excess of GBP 270 million. The annual cash flows attached to that are GBP 160 million. And we're going to go through and show the bridge from EBITDA to cash flow, highlighting things like lease costs that are in there, highlighting working capital movements that are in there, so you can clearly understand that dynamic. Ocado Retail under that model will be a self-financing business in the midterm and will generate annual cash flows in the midterm sufficient to fund growth in excess of 20% of that. And what we mean by that, we'll see the details shortly, is that the cash flows ex any growth CapEx investment is sufficient to invest in modules that will generate over 20% growth on the existing module base at the end of that year. That makes sense. So it becomes a self-financing but also growth business generating capital to invest for growth, really important point. Hopefully, the next few slides will just reinforce those key messages. 2021 and indeed in 2020, Ocado Retail had a very healthy EBITDA margin. One could point that those 2 years were both affected by COVID. Notwithstanding that, we do believe that that sort of margin is obtainable and attainable in the near term. There are several temporary pressures today on margin, the largest of which operational leverage is around GBP 60 million of impact in absolute numbers. And you can see the proportion there of around 90% will serve the margin pressure from operational leverage. Essentially, that's down to the surplus capacity in the CFCs. You've seen at the half year results. We've got -- we're doing around 360,000 orders a week with 600,000 -- with capacity of 600,000 orders per week, so operating at about 60% of capacity. So that capacity to grow into. There are fixed costs there, but also there is -- a portion of those are the Ocado Smart Platform fees that are being incurred and being paid to Ocado Group but part of Ocado Retail's cost base. You've then got marketing costs to grow into that surplus volume. There's around GBP 35 million in that number. And then you've got GBP 25 million of inflation costs around electricity, diesel and dry ice. On a full year basis, those are the 3 absolute numbers. We do expect to be able to address this one, clearly, as we grow into the capacity. Marketing costs are at a point in time to grow into that surplus capacity and inflation, we think there are extraneous factors, the government help in the winter is going to be a good thing to our fiscal '23 numbers, but let's see what happens beyond that. But we do think, ultimately, this is a situation that will normalize. Anything else to mention there. I think that does the job. Re:Imagined will improve the margin profile even further. We expect to head towards that 300 units per hour by the midterm, 275 plus at that point in time, right now Purfleet doing over 200 units per hour and improving. Those are the -- that's the sort of productivity level that we aspire to. On-grid robotic pick going live in '23, '24 will only further improve that productivity. So this higher operational efficiency from increased automation that's in the 7% number, that's around 1% on top of the 6% from On-grid robotic pick alone that increased automation. These benefits will follow on new sites. So in Ocado Retail today, they're not in the Luton site, which goes live in '23, but they will be in future sites going forward. First of all, a reduction of 20% in lease costs and bill savings from the smaller footprint for the same capacity, really important point. And then construction cost savings due to the reduced spec and site size. The size is lower but also the whole infrastructure as well can be smaller under Re:Imagined because you've got that improved productivity and throughput. The new sites will reflect those improvements. The operating benefits retrofittable to live sites, the build benefits will be available on sites going live from '24 onwards. So if we were to take our finger off the pause button to the Northwest and the Southeast, those would follow with those benefits as well. Ocado Retail, the capacity investments that we've already made, cash has already been spent, underpin the revenue growth to GBP 3.9 billion. We're showing here the number of live sites at each of the 3 fiscal years, Luton going live next year, our fourth Zoom going live next year. I think it's the fifth if you include Dordon as well. And the important point is that those sites, as you see here, there's 12 that will be live by the end of '23 will generate revenues of GBP 3.9 billion, again, reinforcing the point around the Northwest and the Southeast sites being paused. The part of the midterm target, driven by 3 key drivers: customer, acquisition and retention. We continue to acquire customers. Average eaches shopped per basket, right now, they're at lower than what might call normalized levels as we -- as the U.K. consumer is operating in the tough market conditions. We're all familiar with and then by a growth in average sales per each. Should reinforce by the way, as a note down here, 1 module is equivalent to GBP 70 million of sales capacity. That includes the impact of inflation. Clearly, under our business model, through of our U.K. partners and our international partners, in January of each year, prices are increased by locals CPI, we've excluded that from these numbers. Ocado Retail, bridging the EBITDA to net cash flow. So how do we get from that GBP 270 million down to GBP 160 million. First of all, Ocado retail incurs maintenance CapEx. This is across the CFCs, across the spokes. It's got IT costs in here. This is general merchandise maintenance CapEx, put all that together, a steady run rate that we've seen over the last few years that we expect to continue of around GBP 15 million. Working capital, as a growing business, we would expect a working capital benefit for Ocado Retail. We received the cash from the customers pretty much instantly. We have 30 days payment terms on average with our suppliers. You put all that together, take into account inventory durations, grocery durations in the grid, we're still delivering a working capital benefit of around GBP 15 million. Lease costs. We have around GBP 65 million of annual lease costs, and that's a pretty good run rate over the next 5 years. That includes rents and capital recharges within that number for the property costs. The property costs vary from CFCs, through to spokes, through to general merchandising sites, through to offices as well. We put in here -- sorry, and the van fleet, of course. We put in here the interest costs that will depend, of course, on any third-party financing that Ocado Retail puts in place. It has been in recent years a self-financing business, but you can very clearly see the disconnects between the EBITDA number and the net cash flow. This year is a cash outflow year for Ocado Retail, and it's funding, I'm sure this will be a question that will come up afterwards is very much dependent on, first of all, growing to start generating cash again. But secondly, it has the resources available to do it today through the revolving credit facility and also the terms that they have with their 2 shareholders, more on that to follow. That net cash flow of GBP 160 million per annum is sufficient to deliver greater than a 20% growth rate, assuming that that capital is invested in those modules that I described earlier. So at that sort of level, Ocado Retail is a cash flow positive business, pre the growth CapEx, but can use that to invest in 20% per annum future growth. Ocado Logistics. So my simple way of thinking about this business, key messages. It's a reliable underlying EBITDA generation business, largely through the nature of its business model. All of its costs are recharged to clients, over Ocado Retail or Morrisons. And then it's 4% management fee on top of that, which is growing at around 10% as volume growth. It's got stable cash flow of around GBP 40 million. There's 2 key drivers that take you from the GBP 30 million to the GBP 40 million. EBITDA growth as volume grows, but offset by the decline in capital recharges by 50% in the midterm. The capital recharges from logistics to its partners today, our round numbers around GBP 20 million, that's going to decline by about GBP 10 million over the next 4 to 6 years as the sites, the Ocado Group sites get fully depreciated. The CapEx is recharged to partners. So again, there's no cash impact on Ocado Logistics. Here's the net -- here's the positive net cash flow model for Ocado Logistics. These were the numbers that we showed back in May of this year, around the number of eaches that are processed, cost recharges and then the revenue that follows EBITDA, mentioning the cash benefit of the capital recharges that we add back to our logistics business, get us to a GBP 40 million positive cash flow for Ocado Logistics. I talked through this already. I've talked to that dynamic. And then it does incur some CapEx which is split as 30%. These are the capital recharges. That capital recharge income is 30% above EBITDA, so included within the EBITDA number. For shared sites, so this is Dordon and Erith in particular. And then for the exclusive sites, the capital recharge income is below EBITDA for those exclusive-use sites. Each of those will fall by around half in the midterm, as I just highlighted, as those assets move closer towards being fully depreciated. Putting that all together, GBP 900 million of revenue in the midterm. That's the 10% per annum growth from today's base. EBITDA, GBP 35 million. The growth in volumes, as I mentioned earlier, that 10% compound annual growth rate will be offset by the capital recharge decline on the shared sites because that's the shared sites sort [ or debt figure ] are in the EBITDA number. This number is outside of EBITDA. That's going to reduce by half from the GBP 14 million to around GBP 7 million, gets you to your net cash flow of greater than 4% of revenue of GBP 40 million per annum. Broadly stable net cash flow as EBITDA growth due to growing volumes is offset by those reducing capital recharges. So the 2 relatively straightforward businesses first and now the one with more detail around it and Ocado Technology Solutions where the value growth opportunity is going to come from for Ocado Group over the midterm as we ramp into the automated warehouse, committed contracts and pipeline of warehouse build that we have ahead of us. Okay. So Technology Solutions. I'll just let you read through these words here. The Ocado Smart Platform, managed service now for 12 global retail partners. The CFCs will be demonstrating this in these materials, provide attractive returns and cash flows, and Ocado Re:Imagined will only improve those further. Combined with the strong management of overheads, and this is an important point that I'll be elaborating on in later slides. We have a component of our cash flows for Technology Solutions around group support costs but around a number of other areas, including board costs and lease costs that I believe we can better manage, keep at least flat in real terms if not reduce from their current levels. At the same time, we have further cash opportunities beyond the core model of the grocery sector for Ocado Technology Solutions. We have with some working capital benefits. We have non-grocery opportunities. We allude to those benefits in these materials, but we take no value for any numbers when we get to our overall summary. It's their all potential upside. So the key messages for Ocado Technology Solutions that you're going to hear about over the next 20 minutes or so. The unit economics of our CFCs continue to improve. We'll go through the detail of each of those, driving improved return on capital employed. With strong cash conversion, we'll go into a bit more detail around those cash outflows and then cash inflows, therefore, a quicker payback. The rollout continues 64 committed sites. It was 58 previously, the 6 that we announced from South Korea a couple of weeks ago takes us to 64. At the end of '22, we'll have 19 CFCs and 4 Zooms, very much on track with that sort of rough guide of 10, 5 module sites per annum as we head towards that 64 number and in excess of 300 modules. There is a clear road map through to the midterm of annual cash flows of GBP 350 million for this business ex growth CapEx. But again, that growth CapEx allows the business to continue to grow if we were on a cost-neutral basis. On that point, the current trajectory will be sufficient to self-fund compound annual growth rate in excess of 20% by the midterm. So the inflows and the outflows. The inflows for Ocado Technology Solutions related to the upfront fees, 4% sales from our partners and the recurring fees related to a number of modules have drawn capacity less the direct operating costs. I'm going to go into the unit economics of each of those 2 final items there on the inflows in a couple of slides. New CFC CapEx is the cash invested upfront in rolling out then the additional modules to drive future revenues. So 2 distinct phases: R&D and overheads and here are the 3 chunks of cash outflow in Technology Solutions. R&D and again using round numbers, you'll see the details shortly and the items that make up those numbers. Cash invested in technology around GBP 200 million. Cash invested in overheads, group support costs to run the business around GBP 100 million and then a bunch of other items, and we'll go through those details again around another GBP 100 million of cash costs. That's the GBP 400 million of cash cost here. And then clearly, cash from partners being the driving key inflows over here. There's your growth CapEx and then there's your cash cost of GBP 400 million per annum in the midterm for the Ocado Tech Solutions business. So let's just look at the CFC component. Here are the cash inflows. We've used year 0 as go-live. And you can see here the upfront fees that we get when the client signs their first order, we immediately receive cash before we've incurred any CapEx. First, CapEx or cash outflows that we start to incur are a very small sliver of ongoing operating costs as we set up a local engineering team. We call that our client service team. We've got a bit of a glossary in a second that explains the various terminology for how we describe our teams. And then the CapEx investment is a combination of the -- well, the fixed assets, the grid infrastructure and so on. The variable assets are the bots and the peripherals. This is the fixed asset investment, the heavy part of it in year minus 1 and 0. And as we add modules, that's when you can see that the cash outflows are incurred then. The ongoing operating costs will grow to a flat level to run the warehouse. The engineering team is typically between 40 to 50 people per site to operate the CFC. We expect to bring that number down. And in fact, it's a reducing number as we'll see shortly, when we highlight what we're seeing at the moment around direct operating costs to run the business. The Re:Imagined technologies underpin the growth to the 40% plus return on capital employed. We're seeing reduced CapEx investment with a reduced initial outflow and higher recurring cash flows, returns per module. The gross CapEx previously, we've guided to 14% of sales. Today, we're indicating that we expect to go below 12%. That is an ambition -- sorry, there is ambition beyond that number. We think we can take it even lower. The mix there is fixed, it's 5%. We brought that down to 4%. Primarily, that's the grid and the infrastructure, again, around those Re:Imagined benefits largely driven by the lighter bots, 600 bots. And then at the same time, the bots and the peripherals are getting less expensive as well, 9% going to 8%. This means that the fixed capital investment going back to my earlier slide, after the 4% contribution from clients is now 0. So we have net zero cash outflow for that initial fixed asset components about 5% out of the 14%, around 1/3 or so is on a cash-neutral basis. Very important when it comes to payback and management of cash flows. Also very important when it comes to we're thinking about financing and peak CapEx and how CapEx should trend over time. The recurring fees, 5%. We're now moving that towards 5.5% as our clients are ready to receive that higher fee in return for even higher returns from the operating productivity improvements that they're going to see. Direct operating costs. We've guided to 1.5%. We didn't have a specific number today, but we expect to move this lower than 1.5%. We'll update you as we go on and get more specific on that number. All current and future OSP sites can be retrofitted for automation improvements. The majority of sites going live in '24 and beyond will include all of these benefits. It's an important slide, this one to go into a bit of a forensics on those first 2 years and how our cash flows develop over time from that pre Re:Imagined guidance of GBP 30 million peak outflow to the GBP 15 million. Cash comes in, starts to incur cost, that's our peak and then we start to grow. Full capacity will here come and is on the next slide, we'll look at the module rollout and then how that cash flow develops further. Qualification on the right-hand side, the exact cash flows will vary, of course, with the capacity ramp because our clients pay for live installed modules. The pace of that ramp will depend on the success of those early modules and filling those early modules. So there's going to be variability there. Clearly, CapEx costs, operating costs, anything local requirements, for example, around a particular CFC, for example, in Japan, where there's a seismic earthquake risk as a different type of warehouse from the ones that we built around the rest of the world. Local inflation. I mentioned earlier, the local inflation will impact fees. Right now, we're seeing elevated inflation levels, fee amounts in absolute terms will grow as a consequence. I think this slide reflects the learnings that we've had from the early international rollout. We've got a much better understanding of the flow of our cash flows went to place orders for CapEx and the management around that. If anything, there is -- that gives a near-term cash flow opportunity as we almost certainly have some surplus CapEx capacity already spent today on our balance sheet for rollouts in '23 and '24. Largely around our better planning of it but also around some of the safety stock that we built up during the supply chain challenges over the last 12, 18 months. So that's a good guy already invested today on our balance sheet. Re:Imagined CFC, taking it out even further. From that peak point again, growing as we go along. The payback is less than 2 years on the module CapEx as we roll that out. The net outflow for fixed cost build is 0, that's just reinforcing my earlier message. Key points on the right-hand side, no CapEx is spent until after the client secures their site. It's a really important point, something we have improved at over time, 0 net cash outflow. We've gone through that model and then the less than 2-year payback for each go-live models. And that's when you start moving to a net outflow state, as you're generating those future models, until you reach a critical mass, even investing in new modules, you're cash flow neutral. So Re:Imagined just underpins that mid-term trajectory. The slide on the left here is an interesting one because it shows the trade-off between higher activity, higher build activity and future returns. So the gray line, for example, where you're doing 15 sites per annum, 75 modules, you'll have a requirement for more CapEx and cash flow in the near term, but the returns and the net present value and the internal rate of return will all be higher as we get into those outer years because you've got a larger base of cash generation. That makes sense. Similarly, if you're on 5 sites per annum, you'll have a cash flow benefit in the near term, but lower returns in terms of cumulative cash flows on the midterm horizon. This little example here shows on 300 modules at year-end and a GBP 400 million total cash base which is the 200 R&D, the 100 group support and the 100 other, how the returns or the amount that's implied to for future growth changes according to the total cost base and the live modules. 26% is that midpoint of 400 costs, 300 live modules at a year-end. But our capability to fund a future growth rate will vary according to the number of live modules at a period end and our ability to bring down that cash cost base for the year proceeding and going forward. That makes sense. So the committed capacity growth to drive significant ramp in cash flows. This exists. We're on track with the plan that we highlighted in May in terms of sites and modules, really key performance indicators for Ocado Technology Solutions around number of live modules, average live modules, capital intensity, again, a really important KPI, all of which will be on a go-forward basis, including as we go through and put this business on the new business model segmentation that I described earlier. Our focus on supporting the client ramp to successfully deliver midterm revenue. We're going to come to that shortly, but it's very much an important driver, of course, to the CFC and module development through to the midterm. Putting this model together, there's a clear path to revenue in excess of GBP 1.1 billion in recurring fees, and that has a greater than 70% contribution margin. We've got recurring fees of 5.5%. We've got direct operating cost of 1.5%. So that 4% return on that 5.5%, very healthy contribution margin to cover those costs of GBP 200 million R&D, GBP 100 million group support, GBP 100 million other costs. And funding those previous growth levels that I highlighted, that little matrix on the previous slide. R&D and overheads. First of all, technology, reaffirming the guidance that we gave in May around the midterm technology cash spend of GBP 200 million. GBP 255 million in fiscal '21. Through to the mid-term, it will decline down to GBP 200 million. We've highlighted the components of that spend. We get a lot of scrutiny on this spend, lot of questions around the returns on capital that we're targeting. I think we've gone through those and what we're aiming for, 22% with Purfleet, 30% plus going forward. Those are the returns we're aiming for from this spend. Radical change portion is something that's nothing like what we're doing today. That could be a Re:Imagined part 2, for example. Innovation for customer needs. We do respond quite a lot to customer requests around their e-commerce model, for example. And so we're very ready to invest for particular customers and any unique features that they want in their model. And then finally, keeping it current around software, hardware, bug fixes, IT security and so on. We'll go through that detail shortly. Any further growth CapEx beyond that GBP 200 million will be to drive step-change innovations and to deliver growth opportunities, that might mean accelerated growth with current and new grocery partners. The non-grocery sector. I expect that with our full year results at the end of February, we'll be able to share some news with you around the business model, potential customers, perhaps even we'll have had some success signing up customers by that point, and we'll go into more detail around our ambitions for our non-grocery sector. In fact, I think generally, the investment case for Ocado whilst there's a midterm perspective here, the world of automated warehouses and the role that we play in that will dwarf, I believe, ultimately, the 64 committed sites that we have today. As a longer-term perspective, we'll get to that one day. And then at the same time, R&D by any further margin enhancements delivered by further automation of CFCs. We've announced On-grid robotic pick. We've announced auto frameload. The next big challenge is the decanting process, taking the goods out from the cardboard boxes that they're arriving from food suppliers, grocery suppliers and putting them in the grid. That's the most difficult part, most difficult manual part of the process today for us to solve. All of this will be self-financed by growth and returns. Any spend beyond that GBP 200 million. I should highlight that tech investment will increase in the near term before heading towards that mid-term target, largely connected to the rollout in fiscal '23 of the Re:Imagined technologies. Group support costs. This is where my own team comes into play here in group operations functions, people, the HR team, the legal team, the finance team and property costs, the gray chunk of spend down at the bottom of the year. And fiscal '21, we were spending GBP 89 million in total across these areas. Then client services and platform implementation. Client services, as a reminder, is the engineering team supporting the international warehouses, keeping the lights on, replacing parts that need replacing, repairing parts that need repairing, and they're part of that components to direct operating costs along with cloud costs. There's a slither of other, which is kindred, as well as other venture-related costs. And then we have our solutions team, which is the sales team, the account management team and then the central client success teams. And I'm going to talk in a second around the client success teams, which is an important new initiative for us at Ocado Group, relatively new, been in place for a few months now, but it's a growing activity. So the client success teams. We've been growing those recently to support our clients in their early ramp and operations. What has become clear is that we can help optimize the international CFCs from the experience that we have learned over 20 years of running warehouses, running logistics, optimizing route planning, managing an e-commerce platform that our partners are all learning those skills from a warehouse perspective. The client success teams is established in 3 regions: North America, in Europe and then APAC, specialist teams, and these are the skills that they're particularly focused on. Helping our clients go faster. There's a rough rule of thumb, 10 to 20 people per partner. A terrific return on capital because if it delivers superior returns for our partners, faster ramp rates, then they clearly -- their orders of modules, CFCs and so on will only grow, is self-fulfilling investment, a very good return on capital. The goal is to help our client get on the road map to optimize the capacity and speed to those CFC economics that we've delivered in the U.K., process through collaboration with our partners, working very, very closely with them. And quite a few examples of tricks of the trade that we've learned over the years that we're now helping our partners deploy in their own models. Look, we've done deep dive comparisons of real-time partner costs of their sites, operational data and compared it to our own and identified how we can get them in line with the sort of metrics that we're used to delivering. And often, it's a short-term deployment as well. This is not necessarily a fixed cost. It's a cost that moves, is transportable, can move around the world, can in fact be remotely monitored in many cases as well, the tools that we now have in place. There's the outcome. We have an agreed road map with our client to support that road map to an optimized performance. So we expect the early investment we're putting in here to drive significant value for the business in the long term, a great return on capital employed for this particular initiative. Other costs, we expect to hold steady at around GBP 100 million. And here's that component of other costs. We've got other CapEx. This is classic sort of office-type IT upgrades. It's certain pre-go-live costs that don't fall within the CFC model. One-off retrofit that bespoke of a particular nature and there might be some innovation enabled in the supply chain. Interest costs, I think that will very -- that line will very much depend on the nature of the financing that we put in place on a go-forward basis. But for the purposes of this exercise, GBP 27 million is a rough guide. Board and other costs, there are 2 gross numbers in here. What I mean by that is the net of that smaller number rather than being gross numbers. The board costs are the nonexecutive costs and the Executive Director costs, offset by the R&D credits that we get for the technology that we're investing in the U.K. to develop the platform and have a small outflow on the lease on the Dordon assets. It will move to 0 in the near term. And then we have various office leases. And we've had a little bit of an uplift. You shouldn't regard the 3D printers as being the full explanation of the GBP 7 million increase, but we've now got a number of 3D printers for the 600 bots. We've developed the 600 bots out of our swiftfield development sites, and we have a fleet of 3D printers to help us with that design and testing of the 600 bots and its components. Net cash outflow, putting all of those together is broadly stable. We're guiding to GBP 100 million number for those costs. On the financing, we have several options open to us. I'm sure in the Q&A, I'm going to get 1 or 2 questions around our plans for financing. Our bond maturity is in '25 and so on. We are actively considering a number of options to ensure optimal funding costs, one of which is that we have done a lot of work on is around our CFCs and the attractive financial profile that they represent for those of a big universe of infrastructure investors that are ready to back and securitize a reliable source of cash flow generation, which is what our CFCs are. We've got very predictable demand drivers. It's a stable, growing cash receipts from that model, high-quality customers, it is a very infra-like and asset-based financing model that we can deploy. We are taking a disciplined approach to this opportunity and indeed other opportunities. Maturities are not for some way out. We don't need to rush. We can take advantage. We can choose our timing and I think our timing has been pretty good so far as we thought about financing. The GBP 500 million bond we did in September of last year, was well timed, at a good cost. And similarly, the raise that we do early this year as well. Getting ahead of the game. So at the right time, we have opportunities to pursue when we're ready to. So Technology Solutions bring that all together. That's the contribution from CFCs, GBP 103 million in fiscal '21, growing to over GBP 750 million. That's the return on those 300 modules after the recurring revenue fee and taking into account the direct operating costs. There's the technology cash spend that I ran through, group support costs GBP 100 million; and then the other, around GBP 100 million there as well, which gets you to GBP 350 million of cash flow ex growth CapEx. But when you look at the average cost per module and you look at the 300 modules at the year-end, that allows an implied module growth of ex 60 modules and 20% investment in future growth from this model, a self-financing model with the ability to fund significant future growth. There are opportunities beyond this model. In-store fulfillment, 7 of our partners plan to roll out the model across their sites. It's early days on this, but the plans are very much there to do that. Growth beyond grocery, I've talked about that. That would almost certainly be a different business model with immediacy of gross margin and profits from that unlike the licensed model that we have today with the Ocado Smart Platform, a good counterbalance to that model. And then on working capital, we believe we've got significant opportunities here. I won't go into the detail right now. Maybe I'll come up in Q&A around our payment terms and what we can do to enhance those. But also just generally better management of our balance sheet. I alluded to the potential for CapEx, cash already invested. There's significant opportunities in our working capital as well, primarily around accounts payable, there's some interesting opportunities for us there. They provide additional flexibility to cover the midterm outlook. None of these benefits are included in those numbers that you've seen on the previous slides. Putting that all together, here's the group summary. Ocado Retail, Logistics and Tech Solutions, midterm net cash flows of GBP 200 million for the first 2, GBP 350 million from Tech Solutions, GBP 550 million by the mid-term to fund future growth. The implied growth rate for each for Ocado Retail and Logistics -- sorry, for Technology Solutions, both in excess of 20%. Here's a summary. Hopefully, I've managed to run through all the evidence to back all of those or at least give you the details of the drivers. Clearly, there's much to do, but the road map is there. Clear path to over GBP 550 million in net cash flow available to finance growth for Ocado Group across those 3 business models. Thank you very much. I'm not going to go through the appendix materials, but they are worth a good read. They -- this one summarizes the responsibility of ourselves and our partners. They get asked a lot of questions around this, around who does what. I'd encourage you to go through this detail at your leisure. These are key modeling assumptions that we've used for this model. And again, I'm sure many of you will be using this to populate your own models going forward. So we just provided that detail in there. And that's it. So that's it from me of the presentation. We've left just over 30 minutes for Q&A, which is probably about right, actually. So over to you.

Thanks for that interesting presentation. Just a couple of things I wanted to clarify that you mentioned circa 60 near-term CFCs. I was just wondering, is that assuming one large partner that announced a lot of CFCs, obviously said it's going to be a slow rollout. So does your assumption of 60, does that include continued slow rollout from that partner? And then like increases from other partners? Or how are you getting to that number? And other thing I wanted to ask was for the 19 that are live, would you be able to give us a range of what the largest and smallest -- are they all around the same number of modules or how wide of a difference it is? And then I guess the last thing on that same question, it seems like at some of the CFCs that have been spokes added. And so have the module -- have the modules changed? Or has all that kind of been according to what the original plan was?

Okay. So remind me the first question again, very quickly.

64 CFCs.

64 CFCs. Yes. So 64 CFCs. We've shown that breakdown of those commitments that our clients have made. We've hired a Kroger, represent 20 CFC of those, Aeon represent a further 20 CFC. The Aeon commitment is, in fact, a sales commitment that we've right extrapolated that to a 20 CFC model. And the balance is made up from various published commitments. What we're really saying is that we're expecting that over that midterm horizon, we will get to that 64 model. It might not necessarily that 60 sites, 300 module end game by that midterm horizon. That won't necessarily be linked directly to a plan today that sort of described because they're in those out years, we don't have those specific timings just yet. It's very much linked '23 and '24, there's more visibility around that rollout because the clients are clearly either identifying their site or have already identified the site and signed it, and we're starting to put CapEx down at the moment. So it's very closely linked to the 64. The timing of it isn't necessarily linear. If anything, you're going to see in '23 and '24, a lower level of sites going live and then the pickup in '25 and '26 and '27 to follow, as our clients continue with their commitments. Does that makes sense?

And then the second one was, what's the variability in the size of sites versus that average?

Good question. Not all sites are equal. The one in Paris is particularly large. The early Kroger sites are particularly large as well. There are a number of sites that are 3 module sites. I'm not going to go into sort of specific partners, but they do vary in size. The rough guide though of 10 sites a year, an average of 5 modules per site is a good guide as to what we've experienced in recent years and what we expect to experience in the next few years as well based on the client orders that are out there, particularly across '23, '24 and '25 sites going live. And the next one?

The last one was, does the addition of spoke, [ so the battle of ] spokes changed the modules basically?

No, they don't. The addition of spokes is more around a client perhaps wanting to expand its population reach rather than anything to do with the CFC itself.

Thanks very much, Stephen. Very interesting. Within that, where should we look at within Technology Solutions and for individual partner CFCs? Where should we be looking at maybe maintenance CapEx or refresh CapEx as we think about securing competitive advantage at the CFC level in maybe year '20 or year '30.

Gosh. I think that would be accommodated within that GBP 200 million technology cash spend. We should be able to accommodate within that number.

I will get to the back eventually. Sorry.

Nick Coulter from Citi. If I may, I'll go one by one. On the CFC rollout in the U.K. and the pause, how does that compare to your original M&S kind of agreements in terms of the earn-out, so to speak? Is that -- does that go back to -- revert to the original or what does it leave you?

I won't go into the exact details of the arrangement with M&S because this sort of degrees of sort of granularity around it that are important. As it stands, it doesn't disrupt those commitments that were made in the agreement with Marks & Spencer, okay? So that remains sort of intact and true to the original agreement in 2019.

And the second one, on the surplus inventory point, how much do you have on your balance sheet? Was that perfect -- it's been really difficult to reconcile the capital outflow with the units going down. So the kind of acknowledgment that you have been running ahead is quite helpful, to be honest, it would be great to know quantification, please?

It's a number that is very much work in progress, but it's -- look at my [indiscernible] particular. But this is a number that is, I would say, in excess of GBP 50 million, probably lower than GBP 100 million. It's that sort of range that I think of capital that for good reasons we've invested in ahead of the curve, particularly when supply chain did look challenged. It puts us in a good place, of course, for CapEx in fiscal '23.

Great. And then the last one, I guess, probably a pivotal one, in that you obviously cash flow negative now and you're aiming to be quite nicely free cash flow positive, but you haven't kind of bridged that kind of journey, what is the trajectory? And then what assumptions do you make around refinancing of converts and the like along the way, please?

Right. So first of all, at the '22 year end in February next year, we're going to be giving guidance around cash flow for fiscal '23. So I think that's probably the right time for you then to be able to see the trajectory. Let's just say we're expecting a meaningful improvement in '23 versus '22 levels. The second question was...

Is it linear broadly or...

It's not -- I would say next year, we'll see a pretty significant improvement in cash flow -- in cash outflow, which will be helpful and largely linked around one, that working capital item that I mentioned; two, around the activity; and three, just I think an improved understanding and management of the cash flows around the CapEx build.

Great. And the last one was just around your financing assumptions as much as you can say given the somewhat sensitive.

Yes. Well, again, I mean, we -- the first maturity that we have on horizon is late 2025, December '25 with a GBP 600 million convertible bonds. Whilst I say there's plenty of time still to go, I mean, naturally, I keep a very close eye on those and how they're trading. And what opportunities we might have around those. There's no need to rush to do anything right now nor are we going to leave it so late and sort of it's not going to be sort of late '24, but we start thinking hard about it. I'm thinking hard about it already. I'm not going into the details of the options open to us, but we do have several. I alluded to the infrastructure-type asset financing that we have available. But we are alert to it. And we have several plans that we already suppressed the button on should we choose to, we don't. Right now, I think the debt markets and indeed, the equity markets to an extent are not as attractive as they have been. So -- and given that we have no need to rush, I think the sensible thing is to have our plans all ready to go, and we'll choose the right time.

Xavier from Bank of America. Two, if I may. The first one, with the new generation of robot on the new technology, are you not afraid that some of your partners may postpone and wait for '24 before putting new CFC? And the second one will be, can you tell us about the length of the contract you've got with your partners or how it works longer term?

Okay. Our partner contracts vary by partner. There's no hard rule of thumb around this. You should be thinking sort of, I think, a rough good guide is sort of in that sort of close to 10-year type horizon. But they very much vary by partner. So that's the first answer to that point. And then the other question was, was it just that one question?

Capex...

Will they postpone because of Re:Imagined ordering?

Well, we are not hearing that from our partners today. They're all very excited by the Re:Imagined products, and we're expecting '24 to include those innovations that we've guided today.

It's Marcus Diebel from JPMorgan. Two questions. One is on the fees, the recurring fees. How flexible are you actually on this? From what you can tell us, is there an element that given the higher volume, we might see also fees potentially going down in percentage terms? Or how strict shall we really think about this again over the next couple of years, yes? That's one. And then the second question related to this is with so much focus on refinancing and what you could do to cover the CapEx. Isn't there also a potential idea that you share this CapEx effectively with your partners, i.e., that they bear more of it and bring, therefore, in the long run, the fee rate potentially down, which comes down to financial -- so that would be an easy way to just avoid this discussion as well and have maybe an alternative to go to the market.

I think -- so on the fees one first, I think if anything, fees are going to be an upside for us going forward. The inflation factor that's built into our contracts is a big plus for us at the moment because that inflation CPI in those territories is ahead of our own CapEx inflation. If anything, CapEx cost. So I've just highlighted are going down rather than going up. So that is a good upside for us. On the CapEx sharing our clients taking more of the share. I think you've got the balance about right, actually. Certainly, our partners say that. It is an option open to us. But I think when it comes to -- I quite like the idea of the returns on capital that we can get from our own CapEx rather than dilute that. And I think it's -- I think we're well placed with the model as we currently stand. So yes, those are the answers there.

So just a few questions from my end. So maybe the first one is you mentioned that for some customers, you try to do custom solutions. And I just wanted to understand how you make sure to get remunerated for that. So you don't overengineer something and then get the same 5% fee. And then maybe just pushing on the GBP 600 million convert in 2026. I know you mentioned you can't give a lot of evidence, but just is an equity issuance on the table as well? And then maybe is the sale of M&S JV also on the table because, as you said, majority of the value you see is in the solutions business. So it would make sense if, for example, you're running out of cash for any reason?

I'd argue exactly the opposite on M&S, on the Marks & Spencer and the JV opportunity. Right now, the reduced margins in that business would really not be a sensible time to sell that business. I think when you look at the road map to that revenue growth for Ocado Retail with the kind of CapEx, the capacity we've already invested and the money has already been spent. I think if there were to be a time to think about it, I'm not hinting that there is at all, that would be the time when you're generating that GBP 270 million of EBITDA and with the growth that goes with that having grown into the capacity and out of the inflationary impacts and the operational leverage that we're experiencing in our numbers today. So that is not an option that we are seriously considering when it comes to the financing. And then on sort of one always has to evaluate financing options in the context of a market at a point in time. There's no rush for us to do anything right here right now. These are pretty choppy waters generally in the equity and the debt markets. I think it makes a lot of sense to have our options open to us, but choose our timing carefully, and it strikes me that doing something sooner rather than later is not very sensible right now.

Are you making sure that you get paid for the....

Yes. Sorry, I should have -- yes, every -- lots of discussions. We are -- it's important for us that we get rewarded for client requests. So reasons that, as you wish, as to how we approach the fees discussion with our partners. But we do get them around whether it be the e-commerce model around the customer-facing app and how that links in with the Ocado Smart Platform. We do get those requests, and we expect to be remunerated for them.

Can I ask on nonfood. You kind of hinted a bit or maybe more than a bit that something would be happening there. Can you kind of elaborate on that? It sounds like it was a kind of low CapEx approach as well. And yes. I guess things like the picking arms would be something that could be very interesting for people, for example, perhaps just a bit more things.

They really are. And I think here, the non-grocery model, again, is something that we've done quite a bit of work on in a short space of time since we announced that model and how we can deploy the technologies that we already have into that space. We are in several live discussions with customers already. We've routinely been approached by customers outside of the grocery sector that want to automate their warehouses. Some very close to home and we've resisted that urge because of the focus on grocery, but it's an opportunity for us. And the robotic arms are exactly that we were just commenting earlier on around the capabilities that Haddington has brought to our technologies for a relatively modest investment and a very healthy return on that capital.

And it would be a lower CapEx model, did you say or…

Yes, it will be. It will be a model that, in fact, we're selling kit directly and probably via third-party integrators rather than directly. Not dissimilar to other players in the space. But it's -- I do think it's a nice balance to our own model where the returns are greater, but take later to arrive versus an immediate. It would be nice to have a business that generates positive revenue, positive gross margin, operating profits and cash flows immediately.

Andrew Gwynn from BNP Paribas Exane with [indiscernible]. Three questions. First one is, I suppose a very simple observation here is a 50% return on capital is pretty impressive. Food retail, online grocery is brutal. Is it too much, do you think? I think you can say I can't answer now, and I'll ask the other 2. Second, maybe more of a request, but I do struggle probably, I think, like other people in this room, to understand the pipeline on the module build out, exactly what's coming when. I appreciate you're not going to give us full visibility to 2030, but it certainly giving a bit more of a sort of line of sight about when we get to '30 or '40 or whatever. And then the final point, actually, it's a useful slide you have there. The direct operating cost of 1.5%. I'll use a Tim-ism, I'm not sure that's the phrase, but Tim's described Triggers broom. So effectively, when you come to the CapEx on the CFCs, you never really come and replace Hatfield, but you do come and replace large parts of Hatfield as you go along. So within the 1.5%, my understanding is that's just the bots. Is there anything in there thinking about the sort of fabric of the hive and so forth. Is that maybe in some of the other CapEx figures that you got on the slide pack?

The 1.5% direct operating costs are the engineering teams and the cloud costs to operate the site. They're not -- there's no CapEx in those numbers.

I thought there was a CapEx in there for the bots replacing kind of maintaining the bots.

In terms of spare parts that need replacing, that's all that is.

But when you think about the trigger broom type analogy, which is the high wears out, maybe a bit of the aluminum break, where is that CapEx?

Well, first of all, we're expecting a much longer duration on any sort of replacement of that. I mean that -- our view is that that's something that will be well beyond 10 years of life. So it's not something that's an immediate consideration for us at the moment. That could comfortably go to 20 years.

The return on capital question. Is it too much?

Is it too much? Well, I think it's really largely around -- you're quite right. The retail sector is a margin -- it's a challenging sector for decent margins. What we believe we have though is a model that delivers that road map to that high single-digit EBITDA number. But it's not just the margin, it's the customer experience, it's the service levels, it's the quality of the accuracy of the delivery, the size of the range of stock keeping units that we offer to our clients. The very much personalized approach to their shopping basket experience. We think that is -- that will only drive more customers and revenues and margin for our customers as well. So we think it's a -- we've had -- we deployed a lot of capital in this space, a lot of risk capital in this space as we've developed where we think the returns that we get are appropriate, and we think that our partners are very well rewarded for investing in the module.

And the module visibility, it's not more of a request really but...

I think we can do this on a sort of -- it's best on a year-by-year basis or maybe sort of a -- on a 2-year bit, we tend to get visibility of a build when the partner finds the site. That remains the longest pole in the tent on this one. And then once they found their site, we can then go to work. But it tends to be an 18-month, 24-month visibility before we know specifically what sites are going live and where. And then we can start giving guidance. So I think otherwise, if we try to give specific guidance, we'll be updating it all along for each year. I think the 10 sites going live per annum, 50 modules live per annum is a good guide that I don't expect will stray far from over the next few years.

Diego [indiscernible] I have a question again on the funding alternatives. And specifically on -- you mentioned the borrowing base facility funding. Can you give us a rough idea of what could be the potential debt capacity for each CFC if you have that? And if the current documentation would allow that already.

Okay. They do allow it. And I think if you were to think around sort of each module generates GBP 70 million of sales with a 4% net return on each module. So you can think around how you could securitize that GBP 2.8 million or so per annum cash flow per module. That gives you a rough guide as to what might be financeable for each module and then put 300 modules around it. The solution will almost certainly be partner-specific due to the nature of what the investor is looking for in terms of some sort of the confidence around the individual partner as well and that what their balance sheet looks like.

Stephen, it's Simon Bowler from Numis. You've kind of quite helpful you kind of split out some absolute numbers in terms of the growth CapEx saving that Re:Imagined is bringing. From a P&L or an operating cost perspective, we see you kind of imply towards 50 bps increase in recurring fees. I believe on the ORL slide earlier, you spoke 200 basis points improvement coming through from Re:Imagined. Is that the right way to think about the kind of operating cost benefit that Re:Imagined is unlocking as a cumulative 150 bps, of which 2/3 you're expecting to pass on to you also...

Sorry, I should have emphasized that particular point, you're absolutely right. It's sort of -- so the way we split the pie of benefit, so to speak, from Re:Imagined is 1/3 Ocado, 2/3 of our partners in very simple terms. We think that's a fair split.

Sreedhar Mahamkali from UBS. Maybe 3 questions, please. Just building on that infrastructure financing question. Is that route open now? Or do you need to demonstrate positive EBITDA at the segment level or individual sort of module level, how does that work actually from a time line perspective?

No, I'll answer that one straight away. Though the route is open now because we have, with our U.K. operation, tangible historical evidence of the ability to generate meaningful cash flows from each module in each site. And we have sufficient evidence around certain of our partners who are getting into further maturing into the ramp to give us that confidence again. So that's what opens up those options.

And does that rely on you getting your partner to agree to be able to do that or...

Yes. There will need to be partner cooperation on this particular idea.

Got it. And the second one, just going back to Andrew's question earlier in terms of the pipeline and visibility. Maybe just ask differently, the 4- to 6-year time frame that you're giving us, is that built on contractual obligations of your partners? Or are you making some assumption beyond that?

No. It's built on the contractual commitments that we have around the 64 CFCs. The exact timing on a number of sites per annum will depend, as I mentioned a little while ago on the clients, identifying the location for their sites and finding the site once they've got that before they start their work that we start our work as well.

Okay. The last one is you've introduced 5.5% potential recurring fee up from 5%. Is that a thought? Or have you actually signed anything that kind of is giving you the confidence that...

These are all live discussions, as you might expect, that we'll update you as we go along as to where we got to.

So we have time -- one more.

It's James Lockyer from Peel Hunt. Just I guess following up on that 5.5% question. Taking us from 5% to the 5.5%, is that 0.5% an average across different partners that perhaps are taking different combinations of solutions from you? And so is the option to go above 6% while up above that 5.5% possible, given on the different options that you might get have to...

That is an average that varies very much by client. And I think I'll leave it at that rather than get into difficult waters.

That's fine. And earlier in the presentation, you mentioned the 20% lease saving because of the Re:Imagined. I think in the recent deal, you mentioned multi-story is possible, which arguably implies half the floor space. So I'm wondering what gets into that 20% and why isn't close to 50% savings?

Well, a lot of it is around that you're able to do the same capacity on a smaller site due to the larger throughput that you're getting from the site from the technologies. So that's going to be a big driver of it. The grid, the bots will all be lighter. That's going to have an impact as well on the construction. So those are the key drivers that get us to that saving.

Sorry, I mean, why was it not higher than 20% given that you can do multistory, which in theory halves of the year?

I think multistory is not going to be the norm. I think it's a possibility, but it's not the norm that we developed as part of Re:Imagined.

And just a final question. You've demonstrated savings that you can make to your clients through Re:Imagined, through technology automation, things like that. But given you are creating a very large grocery buying body with an e-commerce platform that runs across multiple clients, are there other buying opportunities that you might have, for example, payment providers, logistic providers that you might be to provide low margins but savings to your clients that way?

Possible, but challenging. I think territories vary by territory. I think getting that sort of crossover across territories could be really quite tricky. So it's not something that's top of our minds.

Great. So I think we'll take a few from online, so about 5 minutes. And just to say the rest of those that we don't get through, we'll get back to you. Okay. So I think to cover off a few of these, people would like more detail on the inflation pass-through arrangements in the Technology Solutions segment.

So local as a rule of thumb, and I'm not going to go client by client as a rule of thumb. Local CPI is applied in January of each year for our partners on the amount that we have charged before the application of an inflation. So there will be some quite sort of inflated increases as we kick into January next year.

And then apart from refinancing existing debt when they fall due, is there a need for further additional funds to meet the existing rollout of CFCs?

No. I guess now that we've answered that, I think, in the fact that we don't expect to have to do any new financing through to turning cash flow positive. The only financing on the horizon is indeed addressing those maturities at the end of '25, the GBP 500 million bond in '26 and those convert maturities in '27.

And then someone just noting on the logistics slide that we've assumed a 10% growth rate. Is that the same as the targeted Ocado Retail growth rate? Or is there a difference?

It's broadly -- again, it's a 10% compound annual growth rate over a period of time. And again, it won't be linear, but yes, the 2 are aligned, and we shouldn't forget that logistics is around Ocado Retail and Morrisons as well.

And then one Technology Solutions is generating GBP 350 million of operating cash, is the intention to fully invest that for growth? Or if you chose to grow at 10% rather than 20%, would that imply GBP 175 million cash flow post CapEx?

I would rather invest it for growth because I think by that point, return on capital employed should be well towards that 50% number. And I think that's the smartest thing we can do with our cash flow.

Yes. And then -- so the Technology Solutions margin. 70% contribution margin, what do we expect that to be in the midterm post the inclusion of R&D overhead and others as we have that?

Okay. Well, I want to think it very carefully about how much I can say about fiscal '23 through at this point, probably looking at [indiscernible] guidance on this one. I think, first of all, a positive EBITDA in the midterm. I think in the very near term, a positive EBITDA number is very attainable, which this business has not previously been. I think through to the midterm, though, yes, the road map is there. We highlighted it in May, and the things changed from that.

Yes. So 50% was the EBITDA margin that we highlighted in May for that business. So these ones, let's see. FX impact is one that a few people have asked. What is the impact of that on CapEx inflows and outflows?

Yes. We have a highly simple but effective FX method, which is basically what we do is built up our liquidity, our cash reserves in our anticipated currency flows over the next 12, 18 months. So it's a pretty simple model that has worked well for us, let's put it that way. When I look at the balances that we have today and the mixture of currencies aligned to our outflows.

And then we have time for 2 more, maybe. The first, these kind of grouped together, you were asking for what is the expected anticipated revenue and range for international CFC when fully ramped and the expected ramp to maturity.

Gosh. Yes. So again, let's choose a 5 module, CFC, GBP 350 million of sales. We're going to get a 4% return on that. So of GBP 740 million -- sorry, that's the GBP 300 million, that's our cash flows. So we'll get to get the sort of return that we just highlighted towards that 70% contribution margin from those -- from that revenue.

And then have we -- on R&D overhead and other costs, I think how will these be split between the 3 business models?

Well, we're really allocating all of those to technology solutions because Logistics and Retail have their own stand-alone businesses now. Retail has been for some time. We've now set up logistics as a stand-alone business within the group. So all that's left is the group. So if we didn't include them in Technology Solutions, they'd be often costs, which we thought is not particularly helpful. So we've included them as part of tech solutions.

And I think just to finalize because a few more people say, have you assumed inflation in these numbers? Or is there upside to that in the contracts?

That's upside.

Okay. And I think we are at 4:31.

Good. Well, I hope you've joined the session, a lot to get through. Thank you for all your questions, and just give me 1 or 2 areas of thought around where to focus on next. Watch out for the next seminar, which we're actually going to run on ESG in the context of our business, but a completely different subject, but it will be an interesting one. So thank you, everyone. Thank you.